help
help > RE: ROI to ROI DoF
Sep 21, 2018 04:09 PM | Chen-Chia Lan
RE: ROI to ROI DoF
Hi Dr. Nieto-Castanon and Conn experts
I would like to express my deepest gratitude to you for providing this wonderful toolbox for us to explore the functional connectivity of the brain.
And I really appreciate if anyone could please kindly help me out with my confusion regarding the DOF of the f-test.
I further try to change the numbers of the ROIs in the define connectivity matrix pulldown menu manually in the results explorer.
I have a total N=49, and only included an [1,-1] between group contrast.
I found out when selecting only two ROIs, it will show a T-test with DOF=n-2.
When selecting three ROIS, it showed an F-test but with only one number of DOF, F(n-3);
when selecting four ROIs, it showed F(n-4);
...
when selecting nine ROIs, F(n-9);
when selecting 10 ROIs,F(n-10);
when selecting 11 ROIs, it showed F(10,38), and it will always be F(10,38) when I further increased the number of ROIS in the defined connectivity matrix.
(I think these 2 DOFs are calculated from: 49/5=9.8, ceil(9.8)=10; and 49-2-10+1=38.)
Therefore, I am wondering whether when the number of possible connections for a given seed in a defined connectivity matrix is less than ceil(subjectNumber/5), it would change the DOF accordingly, but when exceeding this ceil(subjectNumber/5) number, it would alway use these numbers as the DOF? Would anyone please help explain this approach or if the rationale behind this would be too complicated to explain online, is it possible to point to the related reference?
Thank you very much!
Chen-Chia
I would like to express my deepest gratitude to you for providing this wonderful toolbox for us to explore the functional connectivity of the brain.
And I really appreciate if anyone could please kindly help me out with my confusion regarding the DOF of the f-test.
I further try to change the numbers of the ROIs in the define connectivity matrix pulldown menu manually in the results explorer.
I have a total N=49, and only included an [1,-1] between group contrast.
I found out when selecting only two ROIs, it will show a T-test with DOF=n-2.
When selecting three ROIS, it showed an F-test but with only one number of DOF, F(n-3);
when selecting four ROIs, it showed F(n-4);
...
when selecting nine ROIs, F(n-9);
when selecting 10 ROIs,F(n-10);
when selecting 11 ROIs, it showed F(10,38), and it will always be F(10,38) when I further increased the number of ROIS in the defined connectivity matrix.
(I think these 2 DOFs are calculated from: 49/5=9.8, ceil(9.8)=10; and 49-2-10+1=38.)
Therefore, I am wondering whether when the number of possible connections for a given seed in a defined connectivity matrix is less than ceil(subjectNumber/5), it would change the DOF accordingly, but when exceeding this ceil(subjectNumber/5) number, it would alway use these numbers as the DOF? Would anyone please help explain this approach or if the rationale behind this would be too complicated to explain online, is it possible to point to the related reference?
Thank you very much!
Chen-Chia
Threaded View
Title | Author | Date |
---|---|---|
Megan Fitzhugh | Sep 19, 2018 | |
Chen-Chia Lan | Sep 21, 2018 | |
Alfonso Nieto-Castanon | Sep 22, 2018 | |
Chen-Chia Lan | Sep 20, 2018 | |
allison_shapiro | Sep 20, 2018 | |